[llvm] [InstSimplify] Optimize maximumnum and minimumnum (PR #139581)

[Lewis Crawford via llvm-commits]

================
@@ -6725,32 +6735,50 @@ Value *llvm::simplifyBinaryIntrinsic(Intrinsic::ID IID, Type *ReturnType,
       return Op0;
 
     bool PropagateNaN = IID == Intrinsic::minimum || IID == Intrinsic::maximum;
-    bool IsMin = IID == Intrinsic::minimum || IID == Intrinsic::minnum;
-
-    // minnum(X, nan) -> X
-    // maxnum(X, nan) -> X
-    // minimum(X, nan) -> nan
-    // maximum(X, nan) -> nan
-    if (match(Op1, m_NaN()))
-      return PropagateNaN ? propagateNaN(cast<Constant>(Op1)) : Op0;
+    bool PropagateSNaN = IID == Intrinsic::minnum || IID == Intrinsic::maxnum;
+    bool IsMin = IID == Intrinsic::minimum || IID == Intrinsic::minnum ||
+                 IID == Intrinsic::minimumnum;
+
+    // minnum(x, qnan) -> x
+    // maxnum(x, qnan) -> x
+    // minnum(x, snan) -> qnan
+    // maxnum(x, snan) -> qnan
+    // minimum(X, nan) -> qnan
+    // maximum(X, nan) -> qnan
+    if (PropagateSNaN && match(Op1, m_sNaN())) {
+      return propagateNaN(cast<Constant>(Op1));
+    } else if (match(Op1, m_NaN())) {
+      if (PropagateNaN)
+        return propagateNaN(cast<Constant>(Op1));
+      // In cases like mixed <sNaN, qNaN> vectors, avoid the optimization to
+      // allow correct sNaN propagation where necessary.
+      else if (PropagateSNaN && !match(Op1, m_qNaN()))
+        break;
+      else
+        return Op0;
----------------
LewisCrawford wrote:

I've now rewritten the whole thing elementwise to avoid using any pattern-matchers (or calls to propegateNaN). It now allows for combinations of different foldable vector elements, as long as we return either a full LHS non-constant value, or a full constant value based on the const RHS arg (with some NaN quieting applied).

https://github.com/llvm/llvm-project/pull/139581